How to SOLVE market issues with 5.0.xG version (ideas needed)

koxx · Apr 16, 2009

Too much things in the original post "HTC Sapphire/Vodafone Magic G1 Port".

I'd like to focus on solutions to solve the market issues on 5.0.xG versions.

PLEASE : NO TROLL such as "move to 5.0.xH version, it's better, and it works". Move to another post to troll.

We are here to find solution.

Few ideas have been discussed in the original post such as vendink.apk version or build.prop.

Does anybody have more ideas ?

here is a dump of HTTP packets transiting during a Market application request (useful siwtchers)

It was done (and failed) with 5.0.1G

It could be very interesting to view the same exchange with a working Market version such as 5.0.2H or ADP1. Anybody can do the same capture ? I used BackTrak Linux distribution to did that.

We can see an interesting string in the Market HTTP request : "location,1.0,android,android,en_US.5........g...........g:loc/ql...........1.0.Tandroid/tmobile/kila/dream/trout:1.1/PLAT-RC33/126986:user/ota-rel-keys,release-keys.#2:ddGVOXszuvQlZm8K:1TxOlLJHRfpoZ92D*.fr_FR2.....F SFR .(......server.checkin....settings....location....systemupdater....android....providers.subscribedfeeds....providers.settings"

http://koxx3.free.fr/http_android.cap
http://koxx3.free.fr/http_android.txt

No. Time Source Destination Protocol Info
217 4.254998 192.168.0.104 209.85.229.104 HTTP [TCP Previous segment lost] Continuation or non-HTTP traffic

Frame 217 (881 bytes on wire, 881 bytes captured)
IEEE 802.11 Data, Flags: ...P...T
Logical-Link Control
Internet Protocol, Src: 192.168.0.104 (192.168.0.104), Dst: 209.85.229.104 (209.85.229.104)
Transmission Control Protocol, Src Port: 43119 (43119), Dst Port: http (80), Seq: 139, Ack: 1, Len: 797
Source port: 43119 (43119)
Destination port: http (80)
Sequence number: 139 (relative sequence number)
[Next sequence number: 936 (relative sequence number)]
Acknowledgement number: 1 (relative ack number)
Header length: 32 bytes
Flags: 0x18 (PSH, ACK)
Window size: 5840 (scaled)
Checksum: 0xe072 [correct]
Options: (12 bytes)
[SEQ/ACK analysis]
Hypertext Transfer Protocol
Data (797 bytes)
Data: 00020000226C6F636174696F6E2C312E302C616E64726F69...

0000 08 11 2c 00 00 14 bf 26 7c d8 00 18 41 f3 15 41 ..,....&|...A..A
0010 00 14 bf 26 7c d6 f0 0e aa aa 03 00 00 00 08 00 ...&|...........
0020 45 00 03 51 cd c2 40 00 40 06 f2 15 c0 a8 00 68 E..Q..@.@......h
0030 d1 55 e5 68 a8 6f 00 50 92 10 67 a4 c3 69 e7 ee .U.h.o.P..g..i..
0040 80 18 0b 68 e0 72 00 00 01 01 08 0a 00 05 aa b6 ...h.r..........
0050 92 94 e5 d0 00 02 00 00 22 6c 6f 63 61 74 69 6f ........"locatio
0060 6e 2c 31 2e 30 2c 61 6e 64 72 6f 69 64 2c 61 6e n,1.0,android,an
0070 64 72 6f 69 64 2c 65 6e 5f 55 53 87 35 04 ae 15 droid,en_US.5...
0080 0b e4 9a 00 01 67 00 00 02 e4 00 01 00 00 09 00 .....g..........
0090 08 67 3a 6c 6f 63 2f 71 6c 00 00 00 00 02 d2 0a .g:loc/ql.......
00a0 97 01 0a 03 31 2e 30 12 54 61 6e 64 72 6f 69 64 ....1.0.Tandroid
00b0 2f 74 6d 6f 62 69 6c 65 2f 6b 69 6c 61 2f 64 72 /tmobile/kila/dr
00c0 65 61 6d 2f 74 72 6f 75 74 3a 31 2e 31 2f 50 4c eam/trout:1.1/PL
00d0 41 54 2d 52 43 33 33 2f 31 32 36 39 38 36 3a 75 AT-RC33/126986:u
00e0 73 65 72 2f 6f 74 61 2d 72 65 6c 2d 6b 65 79 73 ser/ota-rel-keys
00f0 2c 72 65 6c 65 61 73 65 2d 6b 65 79 73 1a 23 32 ,release-keys.#2
0100 3a 64 64 47 56 4f 58 73 7a 75 76 51 6c 5a 6d 38 :ddGVOXszuvQlZm8
0110 4b 3a 31 54 78 4f 6c 4c 4a 48 52 66 70 6f 5a 39 K:1TxOlLJHRfpoZ9
0120 32 44 2a 05 66 72 5f 46 52 32 0e 08 05 12 05 46 2D*.fr_FR2.....F
0130 20 53 46 52 20 0a 28 d0 01 12 10 0a 0e 73 65 72 SFR .(......ser
0140 76 65 72 2e 63 68 65 63 6b 69 6e 12 0a 0a 08 73 ver.checkin....s
0150 65 74 74 69 6e 67 73 12 0a 0a 08 6c 6f 63 61 74 ettings....locat
0160 69 6f 6e 12 0f 0a 0d 73 79 73 74 65 6d 75 70 64 ion....systemupd
0170 61 74 65 72 12 09 0a 07 61 6e 64 72 6f 69 64 12 ater....android.
0180 1b 0a 19 70 72 6f 76 69 64 65 72 73 2e 73 75 62 ...providers.sub
0190 73 63 72 69 62 65 64 66 65 65 64 73 12 14 0a 12 scribedfeeds....
01a0 70 72 6f 76 69 64 65 72 73 2e 73 65 74 74 69 6e providers.settin
01b0 67 73 22 bc 03 0a ac 01 0a 0e 08 98 2a 10 90 ce gs".........*...
01c0 04 18 0a 20 d0 01 28 08 10 a7 a1 c4 82 8b 24 1a ... ..(.......$.
01d0 1f 08 ff ff ff ff ff ff ff ff ff 01 10 ff ff ff ................
01e0 ff ff ff ff ff ff 01 18 0a 20 d0 01 28 14 40 72 ......... ..(.@r
01f0 1a 20 08 ff ff ff ff ff ff ff ff ff 01 10 ff ff . ..............
0200 ff ff ff ff ff ff ff 01 18 0a 20 d0 01 28 00 40 .......... ..(.@
0210 94 03 22 12 08 98 2a 10 90 ce 04 18 0a 20 d0 01 .."...*...... ..
0220 28 08 30 c7 f4 07 22 12 08 98 2a 10 8e ce 04 18 (.0..."...*.....
0230 0a 20 d0 01 28 09 30 d3 d6 07 22 12 08 98 2a 10 . ..(.0..."...*.
0240 90 ce 04 18 0a 20 d0 01 28 09 30 f2 9e 04 22 12 ..... ..(.0...".
0250 08 98 2a 10 8e ce 04 18 0a 20 d0 01 28 08 30 f6 ..*...... ..(.0.
0260 ba 03 28 02 12 85 02 08 f7 a5 c4 82 8b 24 12 13 ..(..........$..
0270 0a 11 30 30 3a 31 34 3a 62 66 3a 32 36 3a 37 63 ..00:14:bf:26:7c
0280 3a 64 38 12 13 0a 11 30 30 3a 31 61 3a 36 62 3a :d8....00:1a:6b:
0290 62 34 3a 61 34 3a 34 65 12 13 0a 11 30 30 3a 31 b4:a4:4e....00:1
02a0 36 3a 34 31 3a 30 34 3a 32 65 3a 35 65 12 13 0a 6:41:04:2e:5e...
02b0 11 30 30 3a 31 61 3a 32 62 3a 36 63 3a 66 65 3a .00:1a:2b:6c:fe:
02c0 62 34 12 13 0a 11 30 30 3a 30 37 3a 63 62 3a 35 b4....00:07:cb:5
02d0 30 3a 32 31 3a 35 36 12 13 0a 11 30 30 3a 31 36 0:21:56....00:16
02e0 3a 34 31 3a 30 34 3a 36 34 3a 64 61 12 13 0a 11 :41:04:64:da....
02f0 61 32 3a 34 61 3a 35 35 3a 61 33 3a 34 33 3a 66 a2:4a:55:a3:43:f
0300 37 12 13 0a 11 37 65 3a 62 64 3a 66 33 3a 39 35 7....7e:bd:f3:95
0310 3a 64 39 3a 36 37 12 13 0a 11 30 30 3a 31 36 3a :d9:67....00:16:
0320 63 65 3a 32 35 3a 37 36 3a 64 32 12 13 0a 11 30 ce:25:76:d2....0
0330 30 3a 30 33 3a 63 39 3a 36 66 3a 64 38 3a 65 31 0:03:c9:6f:d8:e1
0340 12 13 0a 11 30 30 3a 32 32 3a 33 66 3a 32 65 3a ....00:22:3f:2e:
0350 33 38 3a 35 37 12 13 0a 11 30 30 3a 30 37 3a 63 38:57....00:07:c
0360 62 3a 35 31 3a 39 38 3a 34 37 18 02 9a 06 02 08 b:51:98:47......
0370 01 .

No. Time Source Destination Protocol Info
236 4.440358 209.85.229.104 192.168.0.104 HTTP [TCP Previous segment lost] Continuation or non-HTTP traffic

Frame 236 (419 bytes on wire, 419 bytes captured)
IEEE 802.11 Data, Flags: ..m...F.
Logical-Link Control
Internet Protocol, Src: 209.85.229.104 (209.85.229.104), Dst: 192.168.0.104 (192.168.0.104)
Transmission Control Protocol, Src Port: http (80), Dst Port: 43119 (43119), Seq: 289, Ack: 936, Len: 335
Source port: http (80)
Destination port: 43119 (43119)
Sequence number: 289 (relative sequence number)
[Next sequence number: 624 (relative sequence number)]
Acknowledgement number: 936 (relative ack number)
Header length: 32 bytes
Flags: 0x18 (PSH, ACK)
Window size: 8384 (scaled)
Checksum: 0x10b9 [correct]
Options: (12 bytes)
[SEQ/ACK analysis]
Hypertext Transfer Protocol
Data (335 bytes)
Data: 0002000001478100000900C8000167000001F21F8B080000...

0000 08 22 2c 00 00 18 41 f3 15 41 00 14 bf 26 7c d8 .",...A..A...&|.
0010 00 14 bf 26 7c d6 a0 39 aa aa 03 00 00 00 08 00 ...&|..9........
0020 45 00 01 83 f5 7f 00 00 34 06 18 27 d1 55 e5 68 E.......4..'.U.h
0030 c0 a8 00 68 00 50 a8 6f c3 69 e9 0e 92 10 6a c1 ...h.P.o.i....j.
0040 80 18 00 83 10 b9 00 00 01 01 08 0a 92 94 e6 bb ................
0050 00 05 aa b6 00 02 00 00 01 47 81 00 00 09 00 c8 .........G......
0060 00 01 67 00 00 01 f2 1f 8b 08 00 00 00 00 00 00 ..g.............
0070 00 8d 90 3f 4b c3 50 14 c5 1b eb 10 82 4a 4c 29 ...?K.P......JL)
0080 84 68 21 a8 20 74 7a 79 7f 93 33 b9 09 76 f1 23 .h!. tzy..3..v.#
0090 f8 f2 8f 82 8b ae 3a 08 ae 2e 4e 22 88 f5 03 b8 ......:...N"....
00a0 fa 11 04 c1 41 9c 9c fd 02 6e ee 7a 4b 5b 43 d1 ....A....n.zK[C.
00b0 c1 37 3d ee 39 f7 77 0e d7 6d 05 1f 6d b7 15 84 .7=.9.w..m..m...
00c0 9e b7 fc 76 b8 de eb b2 e7 dc 09 af 9c 78 97 dd ...v.........x..
00d0 3d 3d 9e 5f 6c 45 9b 5e 87 b4 95 21 69 ef 19 69 ==._lE.^...!i..i
00e0 9f ed 78 b0 b3 10 2c b9 d7 7d ff f2 65 31 f4 e2 ..x...,..}..e1..
00f0 57 27 ea 05 6b ee d7 ec 39 7e f3 fd 97 dc 8d 3a W'..k...9~.....:
0100 de 2a 63 48 34 64 02 26 a1 25 4a 1b 6d 7b 3e 25 .*cH4d.&.%J.m{>%
0110 1f 9f 51 f2 f0 86 92 07 f1 7e 63 2c 2a 70 05 a3 ..Q......~c,*p..
0120 51 f2 c9 be e5 90 16 4a c1 0a 48 81 da fc 60 39 Q......J..H...`9
0130 87 a8 c1 2b 88 14 ca 4c b1 e2 94 b0 1b f7 84 3d ...+...L.......=
0140 68 b0 16 3c 87 2e 50 57 c8 65 53 cb 42 e7 34 80 h..<..PW.eS.B.4.
0150 95 90 d5 64 6c c8 51 a2 16 c8 14 ca 0c 7a 86 dd ...dl.Q......z..
0160 1b 63 4f 46 73 6d 99 41 91 43 51 91 04 4a 37 58 .cOFsm.A.CQ..J7X
0170 89 9c 6a 69 98 02 65 fa fb 08 54 58 55 53 ac df ..ji..e...TXUS..
0180 27 ec c3 51 31 87 15 28 28 b9 a6 65 54 c9 d4 c8 '..Q1..((..eT...
0190 c6 f9 b8 fd 2b 3f 41 96 42 9a 6f d5 ef 07 79 f2 ....+?A.B.o...y.
01a0 01 00 00 ...

TemporalShadows · Apr 16, 2009

I've been following the haykuro megathread...so I wanted to throw out what people are talking about there.

It's possible that there is a difference between the way H and G builds are identifying themselves. We've got a dump from the G build, can anyone get one from the H build so we can compare? That's the first step to either identifying this as the problem or ruling it out. If we can rule out the identification factor it may come down to the actual vending.apk just having a bug, or maybe that specific version of vending.apk is being identified by google. Anyway, it's a start.

koxx · Apr 16, 2009

TemporalShadows said:
I've been following the haykuro megathread...so I wanted to throw out what people are talking about there.

It's possible that there is a difference between the way H and G builds are identifying themselves. We've got a dump from the G build, can anyone get one from the H build so we can compare? That's the first step to either identifying this as the problem or ruling it out. If we can rule out the identification factor it may come down to the actual vending.apk just having a bug, or maybe that specific version of vending.apk is being identified by google. Anyway, it's a start.

do you know if anybody try do pick a vendor.apk from H and run it on G ?

another intersting thing can be to explore build.prop (because it is really used for Market exchange) difference to be sure we can't use G market app (with H build.prop file).

can anybody post a /system/build.prop file in H & G versions here ?

dwang · Apr 16, 2009

koxx said:
do you know if anybody try do pick a vendor.apk from H and run it on G ?

another intersting thing can be to explore build.prop (because it is really used for Market exchange) difference to be sure we can't use G market app (with H build.prop file).

can anybody post a /system/build.prop file in H & G versions here ?

H version of vendor.apk doesn't work on G. At least when I tried it.

grokdesigns · Apr 16, 2009

Hasn't this already been fixed?
http://code.google.com/p/sapphire-port-dream/downloads/list

jashsu · Apr 16, 2009

koxx said:
another intersting thing can be to explore build.prop (because it is really used for Market exchange) difference to be sure we can't use G market app (with H build.prop file).

can anybody post a /system/build.prop file in H & G versions here ?

There's no need; they are identical. If you don't believe me, open H and G update.zip in an archive manager and look at the build.prop checksums. They are identical.

koxx · Apr 16, 2009

ok ... "good".

here is 5.0.1G build.prop

http://koxx3.free.fr/build.prop.5.0.1G

we really need 5.0.2H build.prop file

do you konw how to get it ?

Gimpeh · Apr 16, 2009

grokdesigns said:
Hasn't this already been fixed?
http://code.google.com/p/sapphire-port-dream/downloads/list

That fix did not resolve the issue.

haykuro · Apr 16, 2009

build.prop from 5.0G and 5.0H are exactly the same..

i'm looking into the issue now. will post back here when it is resolved.

koxx · Apr 16, 2009

common string between build.prop and HTTP request :

ro.build.fingerprint=tmobile/kila/dream/trout:1.1/PLAT-RC33/126986:user/ota-rel-keys,release-keys

HTTP request : "location,1.0,android,android,en_US.5........g.... .......g:loc/ql...........1.0.Tandroid/tmobile/kila/dream/trout:1.1/PLAT-RC33/126986:user/ota-rel-keys,release-keys.#2:ddGVOXszuvQlZm8K:1TxOlLJHRfpoZ92D*.fr_FR2. ....F SFR .(......server.checkin....settings....location.... systemupdater....android....providers.subscribedfe eds....providers.settings"

emms007 · Apr 16, 2009

Hey Koxx; do you have any quick way of getting your tcp-dump data easily? I'd be glad to help then.

koxx · Apr 16, 2009

haykuro said:
build.prop from 5.0G and 5.0H are exactly the same..

OK, no need to look into this point.

haykuro said:
i'm looking into the issue now. will post back here when it is resolved.

Can the "new" G market application transmit more or different informations with same build.prop ?

koxx · Apr 16, 2009

emms007 said:
Hey Koxx; do you have any quick way of getting your tcp-dump data easily? I'd be glad to help then.

easy, not exactly, but not too long.

You need a Laptop with Wifi & a home acces point.
You download a Linux distribution specialized for wifi security test (suck as BackTrak, here : http://www.remote-exploit.org/backtrack_download.html)
You boot on this Live CD.

Here are commands to log wifi packets :
ifconfig wlan0 up
airmon-ng start wlan0
airodump-ng wlan0
airodump-ng wlan0 -c 7 -w nom_fichier # 7 etant le canal wifi des echanges entre ton AP et ton G1

easy

J'avais jamais fait, ca m'a pris moins d'une heure.

TheDudeOfLife · Apr 16, 2009

Typically...if we're talking about the old rules (no tinfoil hats) the build was identified by the market using the ro.build.fingerprint from the build.prop.

That's what we had to change to make PAID apps work in both builds.

BUT, you can copy the H build.prop to G and it doesn't fix the problem.

I have yet to see anywhere what Haykuro did to make the R2 builds.

jashsu · Apr 16, 2009

Try copying H version's gtalkservice.apk along with Vending.apk

jashsu · Apr 16, 2009

TheDudeOfLife said:
BUT, you can copy the H build.prop to G and it doesn't fix the problem.

Did you read what I (and subsequently Hakuro) wrote above? the build properties are identical.

whatchamccallum · Apr 16, 2009

I did a tcpdump on my 5.0.2H and did a successful download. follow the tcpstream in wireshark and I don't see anything similar to

HTTP request : "location,1.0,android,android,en_US.5........g .... .......g:loc/ql...........1.0.Tandroid/tmobile/kila/dream/trout:1.1/PLAT-RC33/126986:user/ota-rel-keys,release-keys.#2:ddGVOXszuvQlZm8K:1TxOlLJHRfpoZ92D*.fr_FR2. ....F SFR .(......server.checkin....settings....location.... systemupdater....android....providers.subscribedfe eds....providers.settings"

mrcrb · Apr 16, 2009

As I said in the other thread I am sure the market issue has something to do with the market application version. With the almost same build.prop between H and G there's no difference exchanging them.

The interesting part is that G says for "build number" (directly on the phone) PLAT-RC33 (or something like that, forgotten) with the same build.prop. So I edited the build.prop many many times and tried much more options and and and.

Yesterday I flashed back to JF. Installed all my apps directly from the market again. This morning I've downloaded a new application and after that I saw the new uploaded Haykuro build which said "fixed market". I've flashed (wiped twice before and once after) and then configured (APN, wireless networks, etc.) and rebooted the device. After the device rebooted I saw the arrow-down-icon in the notification bar - the application I've downloaded this morning was automatically downloaded and installed. I was happy and thought "Cool, it's really fixed!". But: I couldn't download anything form the market (stalled "downloading file").

My guess is, that Google updated something in the communication between server (market online) and client (market application) for the upcoming Android phones and/or SDK 1.5 - so the "old" client versions act like before and the "new" client versions are somehow blocked/stalled/whatever.

koxx · Apr 16, 2009

whatchamccallum said:
I did a tcpdump on my 5.0.2H and did a successful download. follow the tcpstream in wireshark and I don't see anything similar to

perfect !!!!
can you give all the exchange ? with hexa values ?
and the HTTP answer from google server ?

MarcusMaximus04@gmail.com · Apr 16, 2009

mrcrb said:
As I said in the other thread I am sure the market issue has something to do with the market application version. With the almost same build.prop between H and G there's no difference exchanging them.

The interesting part is that G says for "build number" (directly on the phone) PLAT-RC33 (or something like that, forgotten) with the same build.prop. So I edited the build.prop many many times and tried much more options and and and.

Yesterday I flashed back to JF. Installed all my apps directly from the market again. This morning I've downloaded a new application and after that I saw the new uploaded Haykuro build which said "fixed market". I've flashed (wiped twice before and once after) and then configured (APN, wireless networks, etc.) and rebooted the device. After the device rebooted I saw the arrow-down-icon in the notification bar - the application I've downloaded this morning was automatically downloaded and installed. I was happy and thought "Cool, it's really fixed!". But: I couldn't download anything form the market (stalled "downloading file").

My guess is, that Google updated something in the communication between server (market online) and client (market application) for the upcoming Android phones and/or SDK 1.5 - so the "old" client versions act like before and the "new" client versions are somehow blocked/stalled/whatever.

sounds solid, they're probably updating for 1.5 and did something so that only version 1.5 would be effected by the maintenance, thus causing the slowdown for us(my guess this is probably just a different market application, maybe one that goes through a different server). It definitely isn't blocking downloads though as I've had 3 go through so far. Each one just took a ridiculously long time.

How to SOLVE market issues with 5.0.xG version (ideas needed)

Senior Member

Senior Member

Senior Member

Senior Member

Senior Member

Senior Member

Senior Member

Senior Member

Senior Member

Senior Member

Senior Member

Senior Member

Senior Member

Senior Member

Senior Member

Senior Member

Member

Member

Senior Member

Senior Member

Similar threads